Optical printing head systems are known in the art and are currently being used in a variety of applications. One way of constructing optical printing head systems is to use one or more high power laser diode bars (LDB) or laser diode arrays (LDA). However, the optical printing head systems in current use suffer from several disadvantages, discussed below.
A schematic diagram illustrating a conventional laser diode bar (LDB) is shown in FIG. 1A. The LDB 12 houses a plurality of individual laser diode elements 14. For example, each laser diode element has a length L.sub.LD of approximately 50 to 150 microns (50 microns is shown in the Figure), a height H.sub.LD of approximately 1 micron and a pitch P.sub.LD of approximately 120 to 600 microns. The y-axis perpendicular to the length of the LDB is termed the fast axis and the x-axis parallel to the length of the LDB is termed the slow axis. The orientation of the LDB is further defined by the XYZ axis shown to the left of the LDB in FIG. 1A.
The near field image of the light source in the slow axis is illustrated in FIG. 1B. The image is plotted as the intensity versus spatial aperture coordinates. The near field extends up to approximately a few millimeters from the lens. As is apparent from the Figure, near field image in the slow axis has a top hat shape. However, after a few millimeters, the far field loses its top hat shape. The far field image of the light source in the slow axis is illustrated in FIG. 1C which is a plot of energy distribution or intensity versus angular scale for a each of the emitters. The far field illumination in the slow axis at the SLM plane has an undesirable depression that causes the SLM to be illuminated non uniformly.
A prior art spatial light modulator illumination system imaging the far field of the LDB source is shown in FIG. 2A. A linear array laser diode bar (LDB) source 12 comprises a plurality of individual elements represented by rectangles 14. Light emitted from these elements is directed to optical element 16 which may be a cylindrical lens. If lens 16 is placed at a focal length distance from the light source 12, as indicated by the distance `f`, the output of the lens 16 will be a set of collimated beams that overlap in the front focal plane. The output of the lens is incident onto a spatial light modulator (SLM) 18 placed at the focal length `f` of the optical element 16 as indicated in the Figure. The SLM 18 comprises a linear array of elements 20. The SLM functions to turn each individual pixel to be illuminated on and off in accordance with the pattern to be imaged.
Using a linear diode array bar to illuminate an SLM, however, suffers from several disadvantages due to the laser's poor emission characteristics. The most noticeable is the well known `smile` affect caused by misalignment of the individual laser diode elements with each other causing some of the elements to lie off axis, thus resembling a smile. This off axis element causes a significant deterioration of optical performance. As an example, the width of a typical laser diode element is approximately 1 micron. An off axis laser diode element may be off as much as 10 micrometers. This results in an error distance of a few tens of micrometers at the SLM, which is large enough such that the light totally misses the active portion of the SLM 20. This error is due to the fact that the light from the laser diode element is magnified in order to cover the bigger pixel area of an SLM.
A major disadvantage of the optical system of FIG. 2 is the large numerical aperture (NA) in the SLM plane 18. This high NA may cause optical crosstalk between adjacent pixels thus decreasing the functional depth of focus of the SLM image.
With reference to FIGS. 1, 2A and 2B, the following example illustrates the disadvantages and limitations of prior art optical illumination systems in connection with the NA. From FIG. 2A the following expression for .theta..sub.OUT is derived from first order optics principles and trigonometry. ##EQU1## dividing both side by tan(.theta..sub.OUT) results in ##EQU2## where N: is the number of individual elements in the laser diode bar
P.sub.LD : is the pitch distance between two laser emitter elements PA1 .theta..sub.IN : is half the angular opening of each element of the SLM in the slow axis PA1 G.sub.SLM =6 microns PA1 .theta..sub.IN =12 degrees PA1 L.sub.SLM =4 mm
Since the angles involved in this example are fairly high, e.g., greater than 10 degrees, the assumption EQU .theta..apprxeq.tan (.theta.)
cannot be made.
We would like to calculate the maximum depth D.sub.SLM for the SLM using the prior art optical system of FIG. 2A. With reference to FIG. 2B we can state that ##EQU3## For the light to pass through the SLM without crosstalk, for a given D.sub.SLM and .theta..sub.OUT, the depth of the SLM cannot be greater than a certain amount. This equation can be rewritten as ##EQU4## In this example, the following parameters are given: NP.sub.LD =10 mm
Substituting these quantities into the equation above yields ##EQU5## Thus, the maximum useable depth of the SLM in this case is 11 microns. This number is not practical as typical SLMs have a depth six or seven times this length. To achieve a low NA, the focal length of the optical element 16 is made as large as possible. A large focal length, however, makes it more difficult to capture the light from the laser diode emitters. As the angle .theta..sub.OUT decreases, the NA decreases and SLMs with more depth can be utilized. Thus, this example, highlights why optical systems such as the one of FIG. 2A are impractical to be used to illuminate SLMs.
Another major disadvantage of the optical system of FIG. 2A is that the far field pattern of the LDB is illuminated onto SLM rather than the near field. This is not desirable because the illumination is not uniform, exhibiting a hat profile having a deep depression at the top of the hat, similar to that shown in FIG. 1C.
Optical systems for transforming the output of a laser diode array are known in the art. The paper titled Geometrical Transformation of Linear Diode-laser Arrays for Longitudinal Pumping of Solid-State Lasers, by James R. Leger and William C. Goltsos, IEEE Journal on Quantum Electronics, Volume 28, 1992, pages 1088 to 1100, discloses an optical system for geometrically transforming a linear laser diode array into a two dimensional, symmetric virtual source with symmetrical divergence. Application of the system to SLM is not disclosed.
U.S. Pat. No. 5,333,077, issued to Legar et al., teaches a lens system for use with a single light beam having a spatial cross sectional distribution which is extended predominately in one direction. In addition, the invention can be applied to a linear array of light beams with each individual light beam being directed to different locations on an imaginary plane by a first optical element wherein at least two portions of each individual light beam are directed along two non parallel paths so as to generate a two dimensional pattern of light on the imaginary plane. A second optical element located at the imaginary plane and aligned with the two dimensional pattern of light redirects each portion of the two dimensional pattern of light to a focal point.
In addition, the paper entitled High Power Multi-Channel Writing Heads, by Dan Gelbart, discloses several schemes for multi-channel high powered writing heads for processless materials. Several schemes highlighted include contiguous, segmented, interleaved, slanted and two dimensional array techniques.